Analysis of Transfer Mode of Airport Passengers Based on Decision Tree

This paper using a transfer selecting model based on decision tree to analysis the behavior of airport passengers. To improve the quality of sample set extracted from randomly questionnaire of passengers in capital airport, the Bayesian estimation iteration model based on information gain has been used to estimate the missing value of the sample set. After that, the model using the decision tree to predict the anticipant traffic modes of airport passengers. This paper emphatically analyzes the passengers who select the light rail. In order to get main factors of transfer selecting in different passenger groups, four kinds of features related to time cost, age, vocation, and trip purpose of the passengers has been used respectively to divide them who more likely to select light rail as transfer modes. The results of the analysis indicate the effectiveness of the model. It provides new thought of prediction of transfer selecting and theoretical support to the building of light rail.